Hi everyone,

(I hope I'm not breaking rules by posting one thread rather than 3 separate ones...)

- what envelope types can be used to control pitch?  (I am using mxadsr for midi control, but can't get it to control pitch correctly. See csd.file)

- how can I sync two oscillators?

- how do I modulate the pulsewave width?

Thanks in advance for any suggestions! :)

brian

Hi there brian. I'm still no expert, but I'll try and help as much as I can.

1. There are a good number of opcodes that can be used as envelopes. I
haven't done much work with midi, so I can't speak for how well each
works with midi, but here is a list of a few other than the ADSR
family:

linen
linseg
expseg
transeg
envlpx

(all of the above have versions that end with an "r" aswell)

If you have the manual, go to page 155 (at least, if you have the
manual for 6.05)

As for the pitch, what was the exact nature of the problem? And are
you dealing with pitch in hz (cycles per second) notation or one of
the logarithmic notations, ie. pitch or octave?

2. syncphasor can be used for oscillator sync. I havent got time to
give an example now, but if you're not up toscratch on it tomorrow
i'll send you one.

3. This depends how you're generating your pulse wave. If it's in your
example, I'll again have to have a gander tomorrow. Sorry man!

Pete

On Tue, Jan 5, 2016 at 4:47 PM,   wrote:
> Hi everyone,
>
> (I hope I'm not breaking rules by posting one thread rather than 3 separate
> ones...)
>
> - what envelope types can be used to control pitch?  (I am using mxadsr for
> midi control, but can't get it to control pitch correctly. See csd.file)
>
> - how can I sync two oscillators?
>
> - how do I modulate the pulsewave width?
>
>
>
> Thanks in advance for any suggestions! :)
>
> brian
>
>
>
>
> Csound mailing list Csound@listserv.heanet.ie
> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to
> https://github.com/csound/csound/issues Discussions of bugs and features can
> be posted here

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here

Hi sorry, I have infact had a chance to look at your example. I don't
know yet what your problem with the pitch envelope is, but the first
two things that strike me are as follows:

 - The MXADSR opcode produces numbers in the range of 0 to 1.
 - The MXADSR creates straight lines for the rising and falling, which
isn't nessacarily going to sound right when used with pitch.

In your example, where it says:

kadsrpitch    mxadsr        i(gkatt3),i(gkdec3),i(gksus3),i(gkrel3)
klfo    oscili    gklfo,gkrate,1        ;1 is sinewave shape
avco1 vco2    iamp,(ifreq*kadsrpitch)+klfo,10        ;10 is square
avco2 vco2    iamp,((ifreq*kadsrpitch)*gkdet)+klfo,0    ;0 is saw

You are multiplying your frequency by a value from 0 to 1. This means
the envelope will take the frequency from 0 Hz up to ifreq during the
attack phase, decay to some value between 0 Hz and ifreq, sustain at
this new value and then decay back to 0 Hz again. Is this the odd
behaviour you are experiencing?

If so, you will either want to add a base frequency to this aswell, like:

(ifreq*kadsrpitch) + ifreq

If you were to do the above, it would start at ifreq, and rise to
twice ifreq, ie. an octave higher. This may not be the behaviour you
are after, so to achieve your desired behaviour may require some
experiments.

Alternatively, you might consider one of the "segr" family of opcodes:

linsegr
expsegr
transegr

These all allow you to set exact values and times for each section of
the envelope, so you can set the start value to 1, or to 0.5, or
something that more closely matches the behaviour you require. These
versions all also allow a release segment (signified by the "r" at the
end). The difference between each one is that linseg creates linear
lines, expseg creates exponential curved lines, and transeg allows you
to set whether the lines are curved or straight, and how curved they
are.

As we hear pitch logarithmically ( ie, 1 octave up = doubling in
frquency). I would recommend having a go with transeg if you want the
pitch to sound like it rises and falls in a linear fassion, although
that might not be what you're after.

Sorry if this answer was a bit long winded, but I hope it helps, let
me know! I'll get back to you on the other two points tomorrow, unless
someone else has covered it by then.

Pete

On Tue, Jan 5, 2016 at 7:38 PM, Victor Lazzarini
 wrote:
> Exponential envelopes are more suited to
> freq control, expon/expseg/expsegr
>
> hard sync can be constructed with
> syncphasor and a table reader.
>
> pwm is built into vco and vco2. Possibly
> vco2 might be sync-able through phase
> modulation.
>
> Victor Lazzarini
> Dean of Arts, Celtic Studies, and Philosophy
> Maynooth University
> Ireland
>
> On 5 Jan 2016, at 16:47, brian@LINUXSYNTHS.COM wrote:
>
> Hi everyone,
>
> (I hope I'm not breaking rules by posting one thread rather than 3 separate
> ones...)
>
> - what envelope types can be used to control pitch?  (I am using mxadsr for
> midi control, but can't get it to control pitch correctly. See csd.file)
>
> - how can I sync two oscillators?
>
> - how do I modulate the pulsewave width?
>
>
>
> Thanks in advance for any suggestions! :)
>
> brian
>
>
>
>
> Csound mailing list Csound@listserv.heanet.ie
> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to
> https://github.com/csound/csound/issues Discussions of bugs and features can
> be posted here
>
> 
>
> Csound mailing list Csound@listserv.heanet.ie
> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to
> https://github.com/csound/csound/issues Discussions of bugs and features can
> be posted here

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here

Hi Peter and Victor,

much thanks for your input! I'm studying both of your posts and trying things out a bit. My thinking was perhaps a bit "simplistic" in that I thought that "ifreq*kadsr" would take whatever input frequency I give at the piano keyboard and send it through the envelope parameters as-is, kind of like what happens with the amplitude.

I tried adding an additional "ifreq" to the equation as you suggested Pete and although it's a bit closer to what I want, the timbre is also quite a bit different (and I don't think it's due to the envelope alone).

In the meantime, I'll keep playing around with parameters a bit and see what I come up with. It's lots of fun just experimenting! :)

brian

On Tue, 5 Jan 2016 19:51:16 +0000, Peter Burgess wrote:

Hi sorry, I have infact had a chance to look at your example. I don't
know yet what your problem with the pitch envelope is, but the first
two things that strike me are as follows:

 - The MXADSR opcode produces numbers in the range of 0 to 1.
 - The MXADSR creates straight lines for the rising and falling, which
isn't nessacarily going to sound right when used with pitch.

In your example, where it says:

kadsrpitch    mxadsr        i(gkatt3),i(gkdec3),i(gksus3),i(gkrel3)
klfo    oscili    gklfo,gkrate,1        ;1 is sinewave shape
avco1 vco2    iamp,(ifreq*kadsrpitch)+klfo,10        ;10 is square
avco2 vco2    iamp,((ifreq*kadsrpitch)*gkdet)+klfo,0    ;0 is saw

You are multiplying your frequency by a value from 0 to 1. This means
the envelope will take the frequency from 0 Hz up to ifreq during the
attack phase, decay to some value between 0 Hz and ifreq, sustain at
this new value and then decay back to 0 Hz again. Is this the odd
behaviour you are experiencing?

If so, you will either want to add a base frequency to this aswell, like:

(ifreq*kadsrpitch) + ifreq

If you were to do the above, it would start at ifreq, and rise to
twice ifreq, ie. an octave higher. This may not be the behaviour you
are after, so to achieve your desired behaviour may require some
experiments.

Alternatively, you might consider one of the "segr" family of opcodes:

linsegr
expsegr
transegr

These all allow you to set exact values and times for each section of
the envelope, so you can set the start value to 1, or to 0.5, or
something that more closely matches the behaviour you require. These
versions all also allow a release segment (signified by the "r" at the
end). The difference between each one is that linseg creates linear
lines, expseg creates exponential curved lines, and transeg allows you
to set whether the lines are curved or straight, and how curved they
are.

As we hear pitch logarithmically ( ie, 1 octave up = doubling in
frquency). I would recommend having a go with transeg if you want the
pitch to sound like it rises and falls in a linear fassion, although
that might not be what you're after.

Sorry if this answer was a bit long winded, but I hope it helps, let
me know! I'll get back to you on the other two points tomorrow, unless
someone else has covered it by then.

Pete

On Tue, Jan 5, 2016 at 7:38 PM, Victor Lazzarini
<Victor.Lazzarini@nuim.ie> wrote:

Exponential envelopes are more suited to freq control, expon/expseg/expsegr hard sync can be constructed with syncphasor and a table reader. pwm is built into vco and vco2. Possibly vco2 might be sync-able through phase modulation. Victor Lazzarini Dean of Arts, Celtic Studies, and Philosophy Maynooth University Ireland On 5 Jan 2016, at 16:47, brian@LINUXSYNTHS.COM wrote: Hi everyone, (I hope I'm not breaking rules by posting one thread rather than 3 separate ones...) - what envelope types can be used to control pitch? (I am using mxadsr for midi control, but can't get it to control pitch correctly. See csd.file) - how can I sync two oscillators? - how do I modulate the pulsewave width? Thanks in advance for any suggestions! :) brian Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here

I've been doing a bit of dabbling. At the moment, I'm getting a mysterious amplitude increase when I sustain notes for any length of time. I see that shortening the amplitude *decay* rate will slow it down, but I don't understand how the decay is controlling the increase..

I also find that very short note strikes will cause a long pitch fall (longer than the release is requesting)..

I'm providing here the updated instrument.

brian

On Tue, 5 Jan 2016 21:29:33 +0100, brian@LINUXSYNTHS.COM wrote:

Hi Peter and Victor,

much thanks for your input! I'm studying both of your posts and trying things out a bit. My thinking was perhaps a bit "simplistic" in that I thought that "ifreq*kadsr" would take whatever input frequency I give at the piano keyboard and send it through the envelope parameters as-is, kind of like what happens with the amplitude.

I tried adding an additional "ifreq" to the equation as you suggested Pete and although it's a bit closer to what I want, the timbre is also quite a bit different (and I don't think it's due to the envelope alone).

In the meantime, I'll keep playing around with parameters a bit and see what I come up with. It's lots of fun just experimenting! :)

brian

 

On Tue, 5 Jan 2016 19:51:16 +0000, Peter Burgess wrote:

Hi sorry, I have infact had a chance to look at your example. I don't
know yet what your problem with the pitch envelope is, but the first
two things that strike me are as follows:

 - The MXADSR opcode produces numbers in the range of 0 to 1.
 - The MXADSR creates straight lines for the rising and falling, which
isn't nessacarily going to sound right when used with pitch.

In your example, where it says:

kadsrpitch    mxadsr        i(gkatt3),i(gkdec3),i(gksus3),i(gkrel3)
klfo    oscili    gklfo,gkrate,1        ;1 is sinewave shape
avco1 vco2    iamp,(ifreq*kadsrpitch)+klfo,10        ;10 is square
avco2 vco2    iamp,((ifreq*kadsrpitch)*gkdet)+klfo,0    ;0 is saw

You are multiplying your frequency by a value from 0 to 1. This means
the envelope will take the frequency from 0 Hz up to ifreq during the
attack phase, decay to some value between 0 Hz and ifreq, sustain at
this new value and then decay back to 0 Hz again. Is this the odd
behaviour you are experiencing?

If so, you will either want to add a base frequency to this aswell, like:

(ifreq*kadsrpitch) + ifreq

If you were to do the above, it would start at ifreq, and rise to
twice ifreq, ie. an octave higher. This may not be the behaviour you
are after, so to achieve your desired behaviour may require some
experiments.

Alternatively, you might consider one of the "segr" family of opcodes:

linsegr
expsegr
transegr

These all allow you to set exact values and times for each section of
the envelope, so you can set the start value to 1, or to 0.5, or
something that more closely matches the behaviour you require. These
versions all also allow a release segment (signified by the "r" at the
end). The difference between each one is that linseg creates linear
lines, expseg creates exponential curved lines, and transeg allows you
to set whether the lines are curved or straight, and how curved they
are.

As we hear pitch logarithmically ( ie, 1 octave up = doubling in
frquency). I would recommend having a go with transeg if you want the
pitch to sound like it rises and falls in a linear fassion, although
that might not be what you're after.

Sorry if this answer was a bit long winded, but I hope it helps, let
me know! I'll get back to you on the other two points tomorrow, unless
someone else has covered it by then.

Pete

On Tue, Jan 5, 2016 at 7:38 PM, Victor Lazzarini
<Victor.Lazzarini@nuim.ie> wrote:

Exponential envelopes are more suited to freq control, expon/expseg/expsegr hard sync can be constructed with syncphasor and a table reader. pwm is built into vco and vco2. Possibly vco2 might be sync-able through phase modulation. Victor Lazzarini Dean of Arts, Celtic Studies, and Philosophy Maynooth University Ireland On 5 Jan 2016, at 16:47, brian@LINUXSYNTHS.COM wrote: Hi everyone, (I hope I'm not breaking rules by posting one thread rather than 3 separate ones...) - what envelope types can be used to control pitch? (I am using mxadsr for midi control, but can't get it to control pitch correctly. See csd.file) - how can I sync two oscillators? - how do I modulate the pulsewave width? Thanks in advance for any suggestions! :) brian Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here

 

 

Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here

Hmm, I'm not sure exactly what's causing your volume problem. The only
thing I can think of is if your filter envelope (for the butterbp has
a long attack, it might sound like it's getting louder after the amp
envelopes attack, but that's a stab in the dark.

Talking of the butterbp, you might want to add something to the cutoff
frequency there, similar to the oscillator freq, as it's currently
also following a path from 0 to ifreq and back again.

A further thought on the timbre not being right is that in VCO2...

avco2 vco2    iamp,(ifreq*kadsrpitch)+(ifreq*gkdet)+klfo,0    ;0 is saw

...you multiply the second instance of ifreq by the detune, but not
the first. Again, this might be intensional, or I might be thinking
about it wrong, but I have a hunch that....

avco2 vco2    iamp,(ifreq*gkdet*kadsrpitch)+(ifreq*gkdet)+klfo,0    ;0 is saw

...might work better.


Victor: I've assumed all this time that expseg does purely exponential
curves, which would mean that on the decay/release portions the curve
would be backwards for pitch. Is it actually logarithmic on the way
down and so matches the attack curve?

On Tue, Jan 5, 2016 at 9:10 PM,   wrote:
> I've been doing a bit of dabbling. At the moment, I'm getting a mysterious
> amplitude increase when I sustain notes for any length of time. I see that
> shortening the amplitude *decay* rate will slow it down, but I don't
> understand how the decay is controlling the increase..
>
> I also find that very short note strikes will cause a long pitch fall
> (longer than the release is requesting)..
>
> I'm providing here the updated instrument.
>
>
>
> brian
>
>
>
> On Tue, 5 Jan 2016 21:29:33 +0100, brian@LINUXSYNTHS.COM wrote:
>
> Hi Peter and Victor,
>
> much thanks for your input! I'm studying both of your posts and trying
> things out a bit. My thinking was perhaps a bit "simplistic" in that I
> thought that "ifreq*kadsr" would take whatever input frequency I give at the
> piano keyboard and send it through the envelope parameters as-is, kind of
> like what happens with the amplitude.
>
> I tried adding an additional "ifreq" to the equation as you suggested Pete
> and although it's a bit closer to what I want, the timbre is also quite a
> bit different (and I don't think it's due to the envelope alone).
>
> In the meantime, I'll keep playing around with parameters a bit and see what
> I come up with. It's lots of fun just experimenting! :)
>
> brian
>
>
>
> On Tue, 5 Jan 2016 19:51:16 +0000, Peter Burgess wrote:
>
> Hi sorry, I have infact had a chance to look at your example. I don't
> know yet what your problem with the pitch envelope is, but the first
> two things that strike me are as follows:
>
>  - The MXADSR opcode produces numbers in the range of 0 to 1.
>  - The MXADSR creates straight lines for the rising and falling, which
> isn't nessacarily going to sound right when used with pitch.
>
> In your example, where it says:
>
> kadsrpitch    mxadsr        i(gkatt3),i(gkdec3),i(gksus3),i(gkrel3)
> klfo    oscili    gklfo,gkrate,1        ;1 is sinewave shape
> avco1 vco2    iamp,(ifreq*kadsrpitch)+klfo,10        ;10 is square
> avco2 vco2    iamp,((ifreq*kadsrpitch)*gkdet)+klfo,0    ;0 is saw
>
> You are multiplying your frequency by a value from 0 to 1. This means
> the envelope will take the frequency from 0 Hz up to ifreq during the
> attack phase, decay to some value between 0 Hz and ifreq, sustain at
> this new value and then decay back to 0 Hz again. Is this the odd
> behaviour you are experiencing?
>
> If so, you will either want to add a base frequency to this aswell, like:
>
> (ifreq*kadsrpitch) + ifreq
>
> If you were to do the above, it would start at ifreq, and rise to
> twice ifreq, ie. an octave higher. This may not be the behaviour you
> are after, so to achieve your desired behaviour may require some
> experiments.
>
> Alternatively, you might consider one of the "segr" family of opcodes:
>
> linsegr
> expsegr
> transegr
>
> These all allow you to set exact values and times for each section of
> the envelope, so you can set the start value to 1, or to 0.5, or
> something that more closely matches the behaviour you require. These
> versions all also allow a release segment (signified by the "r" at the
> end). The difference between each one is that linseg creates linear
> lines, expseg creates exponential curved lines, and transeg allows you
> to set whether the lines are curved or straight, and how curved they
> are.
>
> As we hear pitch logarithmically ( ie, 1 octave up = doubling in
> frquency). I would recommend having a go with transeg if you want the
> pitch to sound like it rises and falls in a linear fassion, although
> that might not be what you're after.
>
> Sorry if this answer was a bit long winded, but I hope it helps, let
> me know! I'll get back to you on the other two points tomorrow, unless
> someone else has covered it by then.
>
> Pete
>
> On Tue, Jan 5, 2016 at 7:38 PM, Victor Lazzarini
>  wrote:
>
> Exponential envelopes are more suited to freq control, expon/expseg/expsegr
> hard sync can be constructed with syncphasor and a table reader. pwm is
> built into vco and vco2. Possibly vco2 might be sync-able through phase
> modulation. Victor Lazzarini Dean of Arts, Celtic Studies, and Philosophy
> Maynooth University Ireland On 5 Jan 2016, at 16:47, brian@LINUXSYNTHS.COM
> wrote: Hi everyone, (I hope I'm not breaking rules by posting one thread
> rather than 3 separate ones...) - what envelope types can be used to control
> pitch? (I am using mxadsr for midi control, but can't get it to control
> pitch correctly. See csd.file) - how can I sync two oscillators? - how do I
> modulate the pulsewave width? Thanks in advance for any suggestions! :)
> brian Csound mailing list Csound@listserv.heanet.ie
> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to
> https://github.com/csound/csound/issues Discussions of bugs and features can
> be posted here Csound mailing list Csound@listserv.heanet.ie
> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to
> https://github.com/csound/csound/issues Discussions of bugs and features can
> be posted here
>
> Csound mailing list
> Csound@listserv.heanet.ie
> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
> Send bugs reports to
>         https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here
>
>
>
>
>
> Csound mailing list Csound@listserv.heanet.ie
> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to
> https://github.com/csound/csound/issues Discussions of bugs and features can
> be posted here
>
>
>
>
> Csound mailing list Csound@listserv.heanet.ie
> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to
> https://github.com/csound/csound/issues Discussions of bugs and features can
> be posted here

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here